Searcharxiv⌕ Search

arXiv subjects

Xujun Wei

Publications and source records attributed to Xujun Wei.

3 recordsLinked to original sources

DaYu-TC: a global-regional nested AI system for tropical cyclone forecasting

Accurate tropical cyclone intensity forecasts are essential for disaster preparedness, yet persistent intensity underestimation limits the value of artificial intelligence (AI) weather models for predicting these destructive storms. Here we present DaYu-TC, a global-regional nested system that links the preservation of strong-wind and low-pressure signals with high-resolution regional prediction. Its global and regional models share an architecture and extreme-aware training objectives. The regional model combines evolving local states with global forecasts valid at each target time, connecting large-scale environmental information with finer-scale cyclone evolution. In an independent evaluation of 16 western North Pacific tropical cyclones in 2023, the global model retains competitive conventional forecast skill while reducing five-day maximum-wind root-mean-square error (RMSE) by 6-22% relative to other evaluated global AI models. Relative to the global model, regional nesting reduces maximum-wind RMSE by 22.0-33.1% and minimum-pressure RMSE by 28.6-39.8% across 12-120 h. The complete system outperforms concurrent operational high-resolution forecasts from the European Centre for Medium-Range Weather Forecasts for minimum pressure at 12-120 h and maximum wind speed at 48-96 h, with respective RMSE reductions of 26% and 12% at 72 h. Without retraining, the same regional model enhances three other global AI models, reducing overall maximum-wind RMSE by 22.6-28.6%. Five-day forecasts take under one minute on a single GPU. These results establish a transferable multiscale approach to reducing cyclone intensity underestimation, extending advances in AI weather forecasting towards more accurate prediction of high-impact weather events.

physics.ao-ph↗

DaYu: Data-Driven Model for Geostationary Satellite Observed Cloud Images Forecasting

In the past few years, Artificial Intelligence (AI)-based weather forecasting methods have widely demonstrated strong competitiveness among the weather forecasting systems. However, these methods are insufficient for high-spatial-resolution short-term nowcasting within 6 hours, which is crucial for warning short-duration, mesoscale and small-scale weather events. Geostationary satellite remote sensing provides detailed, high spatio-temporal and all-day observations, which can address the above limitations of existing methods. Therefore, this paper proposed an advanced data-driven thermal infrared cloud images forecasting model, "DaYu." Unlike existing data-driven weather forecasting models, DaYu is specifically designed for geostationary satellite observations, with a temporal resolution of 0.5 hours and a spatial resolution of ${0.05}^\circ$ $\times$ ${0.05}^\circ$. DaYu is based on a large-scale transformer architecture, which enables it to capture fine-grained cloud structures and learn fast-changing spatio-temporal evolution features effectively. Moreover, its attention mechanism design achieves a balance in computational complexity, making it practical for applications. DaYu not only achieves accurate forecasts up to 3 hours with a correlation coefficient higher than 0.9, 6 hours higher than 0.8, and 12 hours higher than 0.7, but also detects short-duration, mesoscale, and small-scale weather events with enhanced detail, effectively addressing the shortcomings of existing methods in providing detailed short-term nowcasting within 6 hours. Furthermore, DaYu has significant potential in short-term climate disaster prevention and mitigation.

physics.ao-ph↗

RankDNN: Learning to Rank for Few-shot Learning

This paper introduces a new few-shot learning pipeline that casts relevance ranking for image retrieval as binary ranking relation classification. In comparison to image classification, ranking relation classification is sample efficient and domain agnostic. Besides, it provides a new perspective on few-shot learning and is complementary to state-of-the-art methods. The core component of our deep neural network is a simple MLP, which takes as input an image triplet encoded as the difference between two vector-Kronecker products, and outputs a binary relevance ranking order. The proposed RankMLP can be built on top of any state-of-the-art feature extractors, and our entire deep neural network is called the ranking deep neural network, or RankDNN. Meanwhile, RankDNN can be flexibly fused with other post-processing methods. During the meta test, RankDNN ranks support images according to their similarity with the query samples, and each query sample is assigned the class label of its nearest neighbor. Experiments demonstrate that RankDNN can effectively improve the performance of its baselines based on a variety of backbones and it outperforms previous state-of-the-art algorithms on multiple few-shot learning benchmarks, including miniImageNet, tieredImageNet, Caltech-UCSD Birds, and CIFAR-FS. Furthermore, experiments on the cross-domain challenge demonstrate the superior transferability of RankDNN.The code is available at: https://github.com/guoqianyu-alberta/RankDNN.

cs.CV↗